This competing continuation project will employ the 7-decade longitudinal Terman data to examine relationships between theoretically key behavioral and psychosocial predictors and longevity/cause of death/healthy aging. This interdisciplinary research will test hypothesized relations involving the mechanisms of physical activity patterns, risk- taking, religiosity and contentment, changes in social integration across time, explanatory style and resilience, gender diagnostics, substance use, and facets of personality as they predict health many years later. The design is an archival prospective cohort study, using statistical survival analyses and related regression analyses. Because of the long-term nature and richness of the data, this project can compare competing models, can examine long-term effects and time-related changes, and can compare major causes of death. This is consistent with calls for intensive studies of individual differences in behavior patterns, health and disease. Much of the raw data come from the archive begun in 1921 by L. Terman, plus extensive additional data already collected by this project on date of death, cause of death, smoking, and various indexes of personality and social stability. The subjects are 856 men and 672 women followed from their childhood in the 1920's until the present, the longest continuous cohort study every conducted. Following up on the ongoing research, which uncovered links between major psychosocial patterns earlier in life and subsequent premature death in middle and old age, the present project will employ reliable behavioral patterns and psychosocial indices to compare models containing these and related later-life psychosocial stresses and resources. This project thus aims to provide hard-to-obtain information relevant to understanding the influence of these social, individual, and behavioral factors on longevity and cause of death across the life span.